RFR 8014217: Base64.getXDecoder().wrap(...).read() doesn't throw exception for an incorrect number of padding chars in the final unit

[Xueming Shen]

It has been discussed before that the IOE is preferred in case of the wrapped
decoding stream. The spec of wrap() does mention that "IOE when reading
bytes that can not be decoded".

So personally I feel it has been "covered", but if preferred, it can go further as

"If there is padding character present in the final unit, the correct number of padding
character(s) must be present, otherwise IllegalArgumentException (or IOException in case
of decoding from the wrapped base64 stream), is thrown during decoding."?

-Sherman

On 05/14/2013 12:25 PM, Chris Hegarty wrote:
> Is there a conflict between the spec and implement? IOE versus IAE?
>
> -Chris
>
> On 14 May 2013, at 19:50, Xueming Shen<xueming.shen at oracle.com>  wrote:
>
>> Hi,
>>
>> Please help review the change for handling a "malformed base64 stream" corner case.
>> The latest spec has been updated/clarified as
>>
>> "If there is padding character present in the final unit, the correct number of padding
>> character(s) must be present, otherwise IllegalArgumentException is thrown during
>> decoding."
>>
>> But the stream decoding implementation has not been updated accordingly. The change
>> is to throw the IOE if the padding character "=" is missing from the xx== pattern.
>>
>> http://cr.openjdk.java.net/~sherman/8014217/webrev/
>>
>> Thanks!
>> Sherman